Wrist accelerometers accurately measure gait for frailty monitoring in older adults

A wrist-worn device can be used to accurately measure gait speed in a real-world setting, providing an easy and reliable method of monitoring age-related functional decline, according to a recent study.

“The results of this study demonstrated the feasibility of using a wrist acceleration-based method for long-term real-world monitoring of gait speed in a large cohort of community-dwelling adults,” the researchers said.

The cross-sectional study included 2,809 participants (aged 45–75 years, 1,508 women) who were asked to wear a wrist device for 13 consecutive days for gait measurement. They were instructed to live as usual and were given no additional intervention or supervision. Raw acceleration data were transformed for computer analysis, where gait bouts were detected and speed was estimated.

Ninety-eight percent of participants had viable gait bout data for at least 8 valid days, allowing for weekday and weekend analysis. The wrist-worn device showed that across all age groups and regardless of day participants moved mostly in <30-s bouts of gait, which comprised >60 percent of the total number of gait bouts. [Sci Rep 2021;11:18966]

According to the researchers, such short-duration bouts of gait are indicative of indoor activity. In comparison, >120-second bouts, which are more likely to occur outdoors, were very rarely detected, comprising <10 percent of the total number of gait bouts. An intermediate bout duration of 30–120 seconds was observed around 30 percent of the time and could a mixture of both indoor and outdoor activity.

To assess the value of incorporating gait parameters in the assessment of frailty status, the researchers generated two multivariable logistic regression nested models. The first (model A) included only age, sex, body mass index, and physical activity intensity, while the second (model B) added gait speed metrics.

Model A had an area under the curve (AUC) of 0.763 for predicting frailty status, which improved to a mean of 0.793, 0.793, and 0.790 when gait bouts of <30s, 30–120s, and >120s were incorporated in model B, respectively. The corresponding likelihood ratio values were 71.06, 68.59, and 49.92, all of which were statistically significant (p<0.001 for all), suggesting that model B was substantially better at distinguishing frailty status.

Similarly, the inclusion of gait bout and speed metrics improved the base model’s predictive ability for handgrip strength. These findings were also robust to subsequent sensitivity analyses, which excluded participants with high-speed bouts or focused on those aged ≥65 years.

“Gait speed, recognized as the ‘6th vital sign’ is a reliable and sensitive measure of people’s functional ability, closely associated with well-being, healthy ageing, physical frailty, and survival in older adult populations,” but studies that have assessed this utility of gait have mostly been conducted in a laboratory or clinical setting, the researchers said. [J Geriatr Phys Ther 2009;32:2-5; Phys Ther 2002;82:128-137]

The present study provided evidence that predictive ability of gait for functional status holds true in free-living settings, too, and that wrist-worn devices could allow for easy and accurate monitoring, they added.